提交 9d195a54 编写于 作者: M Marc Zyngier 提交者: Albert ARIBAUD

ARM: HYP/non-sec: remove MIDR check to validate CBAR

Having a form of whitelist to check if we know of a CPU core
and and obtain CBAR is a bit silly.

It doesn't scale (how about A12, A17, as well as other I don't know
about?), and is actually a property of the SoC, not the core.

So either it works and everybody is happy, or it doesn't and
the u-boot port to this SoC is providing the real address via
a configuration option.

The result of the above is that this code doesn't need to exist,
is thus forcefully removed.
Signed-off-by: NMarc Zyngier <marc.zyngier@arm.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>
上级 e771a3d5
......@@ -30,25 +30,8 @@ static unsigned long get_gicd_base_address(void)
#ifdef CONFIG_ARM_GIC_BASE_ADDRESS
return CONFIG_ARM_GIC_BASE_ADDRESS + GIC_DIST_OFFSET;
#else
unsigned midr;
unsigned periphbase;
/* check whether we are an Cortex-A15 or A7.
* The actual HYP switch should work with all CPUs supporting
* the virtualization extension, but we need the GIC address,
* which we know only for sure for those two CPUs.
*/
asm("mrc p15, 0, %0, c0, c0, 0\n" : "=r"(midr));
switch (midr & MIDR_PRIMARY_PART_MASK) {
case MIDR_CORTEX_A9_R0P1:
case MIDR_CORTEX_A15_R0P0:
case MIDR_CORTEX_A7_R0P0:
break;
default:
printf("nonsec: could not determine GIC address.\n");
return -1;
}
/* get the GIC base address from the CBAR register */
asm("mrc p15, 4, %0, c15, c0, 0\n" : "=r" (periphbase));
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册